veille Built for, Public Sector
Vertical

Public sector AI use is growing. Accountability frameworks are now law.

Ontario's Bill 194, the federal TBS Directive on Automated Decision-Making, and Québec's Law 25 create specific compliance obligations for public-sector organizations using AI. Veille provides continuous monitoring and documentation infrastructure to meet them.

Primary frameworksLaw 25, Bill 194, TBS Directive
Bill 194 statusRoyal Assent Nov. 2024, Standards pending
Law 25 statusFully in force since Sept. 2023

Applicable frameworks

The public sector regulatory stack.

Québec, Privacy Law 25 (QC), Fully in force since September 2023
Applies to public sector bodies in Québec (ministères, Crown corporations, agencies) under the Act respecting Access (A-2.1). The primary AI obligations are automated-decision transparency (Art. 65.2), a privacy impact assessment before deployment (Art. 63.5) and before any release of personal information outside Québec (Art. 70.1), and a designated person in charge of access and protection of personal information (Art. 8). These are the public-sector article numbers, distinct from their private-sector equivalents (s. 12.1, 3.3, 17, 3.1 in the Private Sector Act).
Tier 2, Encoded Bill 194 (ON), Technical standards forthcoming
Applies to Ontario public sector organizations. Requires AI accountability frameworks, AI disclosure to affected individuals, and cybersecurity protocols. Technical standards are forthcoming by regulation. Organizations should begin AI inventory now, before standards are set. Retroactive compliance is more expensive than proactive preparation.
Tier 2, Encoded TBS Directive on Automated Decision-Making (Federal)
Applies to federal government departments and agencies. Requires Algorithmic Impact Assessments (AIA) before deploying automated decision systems, peer review, and bias testing. Four-tier classification system with increasing requirements for higher-impact systems. Veille encodes the Directive article by article, 8 obligations covering AIA, transparency notices, decision explanations, bias testing, data governance, peer review, human intervention and recourse.

Use cases

AI systems in the public sector that require monitoring.

SystemFrameworksKey obligations
Benefits eligibility determination Law 25TBS Human review pathway, transparency notice, AIA before deployment
Document and records classification Law 25Bill 194 PIA for personal data processing, AI disclosure under Bill 194
Fraud and anomaly detection Law 25TBS Transparency for benefit suspensions, AIA, human escalation pathway
Public-facing chatbots Law 25Bill 194 AI disclosure obligations, PIA if processing personal data, accountability framework

Accountability

Public sector AI accountability is no longer optional.

Veille provides the documentation infrastructure that public-sector organizations need to demonstrate compliance, to auditors, to legislators, and to the citizens they serve.

Book a call →